China: building on MPLS
In China, 3G licenses were issued to the three full-service telecom providers in January 2009. However, China Mobile has been operating ten pilot TD-SCDMA networks for the past year, comprising some 18,000 Node B's. These provide service to 200,000 customers but have the potential capacity to support up to 9 million subscribers. The Chinese market will see the complete range of 3G technologies deployed, from homegrown TD-SCDMA to EV-DO and W-CDMA variants.
Today, China Mobile is migrating to IP, and it is likely that the 50,000 base stations for the first commercial phase will ultimately use the new MPLS transport network. As such this leaves the Chinese operators in the enviable position of being able to migrate to an all-IP network. A single IP network able to both support the rapidly growing mobile broadband traffic as well as legacy voice and data services will slash costs thereby making all mobile services more profitable. However, such an upgrade is not without a major challenge.
The entirety of the mobile network, from base stations to core network elements, needs to be functioning based on the same clock. Ethernet, however, is inherently asynchronous. Moreover, packet-switched networks introduce delay, while packets can also go astray. When the network becomes out of sync, calls are dropped, handovers are fumbled and mobile operators' traditional voice service quality consequently drops off significantly. Other real-time services such as video are similarly affected.
It is this area that presents the single greatest technical challenge to all-IP networks. The industry is working hard to overcome these impediments by engineering clocking mechanisms in the packet layer and the Ethernet physical layer. Nevertheless, it is an issue that Chinese operators must take into consideration when planning their networks.
India: optimized Abis
Meanwhile, India is currently witnessing the first 3G service soft launches by MTNL and BSNL in a handful of cities with wider deployments expected once the license auctions have been completed later this year. However in India an all-IP network is not an option. The current infrastructure is not ready for such a radical transformation, and the economics dictate a slow evolution. This means that operators will have to deliver profitable 3G services using their existing infrastructure.
Due to the massive sparsely populated regions with limited fixed-line resources, the Indian scenario is to connect the base stations to the core through the existing microwave network. In this case, Abis optimization will be required to reduce cellular transmission bandwidth while maintaining service quality as well as enabling controlled introduction of early 3G services with no additional capital investment.
Abis optimization can reduce the bandwidth required for 2G traffic by more than 50%, freeing up what's left for 3G. Additionally more aggregation sites will need to be deployed throughout the network, thereby allowing the growing data traffic to be multiplexed and delivered to the network core more efficiently without requiring massive investment in infrastructure.
Such technology will also help to provide 3G services in remote regions in China too - it may not deliver the speeds that urban users will receive, but it will start the process of breaking down the digital divide.